This research demonstration project has four interrelated objectives. The first is to develop a New Integrated Assertive Community Treatment Team for Dual Diagnosed Clients. The previous research demonstrated that the assertive community treatment team was an effective method for assisting homeless mentally ill clients. The proposed project will add specific service components aimed at assisting homeless mentally ill clients who also abuse substances. The second objective is to compare the effectiveness of the New Integrated Treatment Model against Assertive Community Treatment only and a Treatment-as-Usual control group. Two hundred and one dual diagnosed homeless individuals will be randomly assigned to either the new integrated team, an assertive community treatment only condition, or a treatment-as-usual control group which will only be given a packet of information on how to access existing treatment programs. Participants will be followed for 30 months. Outcome variables include days in stable housing, psychiatric symptoms, substance abuse, and consumer satisfaction. The third objective is to conduct a process evaluation. A process evaluation will be conducted to determine how the following predictor variables affect client outcomes: 1) developmental vulnerability, for example, family psychiatric history, psychiatric diagnoses, and 2) ascribed status such as gender and race; 3) stressful events; 4) client coping strategies; 5) social support and therapeutic relationships; 6) specific service activities, for example housing assistance, income assistance, supportive services, payee services, substance abuse services, vocational services, and mental health services. The last objective is to conduct a Survival Analysis predicting both exits from homelessness and the achievement of abstinence. Survival analysis will be used to predict exits from homelessness and the achievement of sobriety. The proposed project will improve on previous research by incorporating longitudinal data on both client characteristics and social service system variables. Prior research on these problems has relied almost exclusively on cross-sectional designs and has only examined the impact of client characteristics.